A RANS-based analysis tool for ducted propeller systems in open water condition

The Maritime Research Institute Netherlands (MARIN), formerly known as the Netherlands Ship Model Basin (NSMB), has carried out in the past a lot of experimental (and theoretical) research on ducted propellers. The investigations have provided useful data on the performance characteristics of this type of propulsion unit, which have found worldwide employment (e.g. Wageningen Ka series, 19A duct). However, the experimental investigations have – as a matter of course – been carried out at model scale and consequently the data may suffer from scale effects. For the propeller, these scale effects are generally considered acceptable (i.e. predictable) because of its high rotational speed. For the flow around the duct, however, this is not so obvious. Transition of the flow to a turbulent state may well be delayed, laminar flow separation may occur, which affects the action of the duct in its interaction with the propeller. Repeatedly the question has therefore arisen: are model tests on ducted propeller systems sufficiently representative for the full scale situation?
A (partial) answer to this question may be obtained from the application of Computational Fluid Dynamics (CFD). Indeed some papers on CFD analysis of ducted propellers have already appeared. In particular, a paper by Abdel-Maksoud and Heinke [1] addresses the scale effect issue for ducted propellers. A rotating grid and an earth-fixed grid are combined with a sliding interface technique to analyse the Ka 5-75 propeller in the 19A duct in open water and an increase of the duct thrust with increasing Reynolds number is reported.
In the present paper we propose a simpler computational model by representing the propeller as an actuator disk of finite thickness, while the duct is maintained in its true shape. As will be explained, this gives a substantial reduction of the computational effort, while the open-water performance of the duct can still be evaluated successfully. Modest computational effort allows the tool to be used in the design process, in which several operating conditions have to be evaluated to arrive at a suitable compromise for the duct shape.
The purpose of this paper is to show that the chosen computational model is suitable for the analysis of the open-water performance of ducted propeller systems. After a description of the model, a variety of results is reported to show that most trends in performance, known from experiments, can well be reproduced. Ample use is made of available data for the 19A and 37 ducts. At the end of the paper a section is devoted to the study of scale effects